Accomplishments: The analysis included women ages 35-49 from the National Health and Examination Survey III. Menopausal status was based on months since last period (less than 2, 2 - 12, and more than 12 for pre-, peri-, and post-menopause, respectively). Factors related to menstrually-defined menopausal status were identified with multinomial logistic modeling. Factors related to elevated FSH were identified with binomial logistic regression with FSH > 20 IU/L as the dependent variable. Women with surgical menopause were excluded from the menstrual-cycle based menopausal status analysis. Both modeling strategies adjusted for age, smoking, and unilateral oophorectomy. Higher body mass index ( >= 30 kg/m2 compared with < 25.0 kg/m2) was protective for elevated FSH (odds ratio, OR, 0.6) but not for the menstrual measure of menopause (OR 1.2). Exercise (>= 3 times per week compared with none) was associated with a lower likelihood of being both menstrually and hormonally post-menopausal (ORs 0.3 and 0.6, respectively). Alcohol use also tended to be protective of menstrually-based menopause and increased FSH, but not significantly so. Other factors examined include ethnicity, education, age at menarche, live births, and oral contraceptive use.